Graviton in Production JVMs: ARM64 Is Not a Flag Day, It Is a GC and JNI Audit
AWS Graviton, multi-arch containers, AArch64 JDK builds, and the native libraries that fail first when you leave x86.
Category
49 articles · 174 min of reading
AWS Graviton, multi-arch containers, AArch64 JDK builds, and the native libraries that fail first when you leave x86.
How SecurityFilterChain is assembled, where UsernamePassword and JWT filters sit, and why 'it works in a test slice' still fails behind a gateway.
Advisors around ChatClient, QuestionAnswerAdvisor, and the retrieval details that matter: chunking, filters, and not stuffing the window.
JEP 431: SequencedCollection, SequencedMap, reversed views, and how new addFirst/getLast methods change APIs without a new List subtype zoo.
How HotSpot decides an object never leaves a method, what scalar replacement actually removes, and how to read JFR when allocation still spikes.
Always-on JFR: which event templates belong in prod, how to dump recordings on a stall, and what JFR will not tell you that async-profiler will.
JEP 454: Arena, MemorySegment, Linker, and why bounded lifetimes beat ByteBuffer folklore for off-heap and native calls.
Truffle ASTs, partial evaluation, and Polyglot Context: when embedding JS or Python in the JVM is justified, and what isolation you actually get.
What structured concurrency adds on top of virtual threads: scopes, cancellation that actually propagates, and how it compares to CompletableFuture spaghetti.
OpenJDK's jcstress harness: litmus tests, outcome histograms, and how to catch publication bugs that unit tests will never see.
JSR-133: what synchronization actually publishes, why double-checked locking needed a rewrite, and how to reason without folklore.
Thompson, Farley, Barker, and co. on the Disruptor: ring buffers, sequence barriers, and avoiding locks when microseconds are the product.
A deep dive into Java stack memory, heap memory, metaspace, and how garbage collection actually reclaims objects.
G1, ZGC, and the flags that steal latency: pause goals, heap sizing, and the allocation rate that no collector can hide.
Design one error contract for your whole API using @RestControllerAdvice, ProblemDetail, and validation groups.
What each SOLID principle actually asks of your code, shown with realistic examples instead of Animal-extends-Dog toys.
A measured look at which Spring Boot startup optimizations move the needle in practice, from lazy initialization to Class Data Sharing and beyond.
How Spring Modulith turns package structure into an enforced architectural boundary, catching the coupling that code review alone tends to miss.
What actually changes when you compile Spring Boot to a GraalVM native image, and the reflection and initialization pitfalls that trip up the build.
Comparing MapStruct's generated mappers against hand-written mapping code across boilerplate, performance, and how each handles change over time.
A System.nanoTime() loop around your code is not a benchmark — it's a measurement of the JIT compiler warming up. JMH exists to fix exactly that.
Wildcard generics confuse most developers on sight, but the PECS rule turns a wall of question marks into a mechanical, memorable decision.
HashMap's average O(1) performance depends on assumptions about hash distribution and load factor that are worth understanding before they're violated.
Why the default executors behind @Scheduled and @Async are wrong for production, and how to configure thread pools that actually match your workload.